Robots, Rice, and the Real World Web

Setting the Stage: Why We Care About Physical AI

We opened with a quick recap of why we exist at all. Earlier this year, Nvidia’s Jensen Huang described the trajectory from generative AI to agentic AI, and then to physical AI—AI that understands space and physics and can act in the real world.

Roughly 70% of global GDP is still tied to physical locations and labor, so going from agentic to physical AI effectively triples AI’s addressable market. If you want to build something bigger than OpenAI, you have to play in physical AI.

That’s why, back in 2021, we started building what we now call the real world web: a way for digital devices to browse physical locations the way humans browse websites.

• Not “one big map,” but many domains, like websites running on different servers
• A collaboratively editable 3D view of the world, built with privacy and performance in mind
• Accessible to phones, glasses, robots, smart city systems, and more

Imagine if the internet was just everyone uploading their information to a specific website. No. The internet is many different websites on many different web servers, and discovery and linkage between them. That’s exactly what the real world web is.

Six Layers of Robotics and Where We Sit

We revisited our view of the six necessary layers for robotics:

1. Locomotion – basic movement (walk/drive, turn, avoid obstacles)
2. Manipulation – interacting with objects (arms, grippers, etc.)
3. Perception – spatial semantic perception, knowing what things are and how far away they are
4. Mapping – memory of where things are
5. Positioning – knowing where you are relative to the map
6. Applications – the actual jobs, workflows, and experiences on top

Our focus is squarely on collaborative perception, mapping, and positioning, plus a growing ecosystem of applications.

Phones and glasses are, in Nils's words, “Really just like a robot but with no arms and legs.”

By starting with handhelds and AR smart glasses, we’ve been able to:

• Deploy Cactus, our retail copilot, into 1,000+ locations
• Generate millions in pilot revenue
• Build an open pipe of $150m+ in demand
• Now connect robots and glasses into the same spatial fabric

That groundwork is exactly why we can now do what we showed next.

Demo: Turning a Brand-New Utility Robot Into a Store Worker in Two Minutes

We showed a demo of a fresh-out-of-the-box utility robot with almost no prep work:

• The venue was already mapped with phones and connected to the real world web.
• The robot was brand new; the only thing we’d done was plug in a USB stick with our app.

Here’s the flow:

1. Phil opens the APK from the USB stick and installs it.
2. On first launch, the robot checks for internet and asks: “Am I plugged into my charging station?”
3. It then requests: “Please take a picture of the QR code on the charger and show it to the camera.”
4. Phil scans the QR code on the charger with his phone and presents it to the robot’s camera.
5. The robot now knows which mapped QR code is its charger.
6. It then pulls the map, application logic, and product locations from our store domain via the network.

Less than two minutes of setup, and this robot is good to go. No field engineer crawling around with a laptop. No tedious one-off mapping. Just “plug in, show QR, done.”

This is one of the core reasons we’re building the real world web: so robots can arrive in new places and instantly know where they are and what to do.

Enter Rice AI: Our New Neighbors and Robotics OGs

Then we handed the mic to Victor and Aaron from Rice AI, our new neighbors in the Level 10 research center in Kowloon Bay, Hong Kong.

Rice in a Nutshell

• Rice has been in robotics since 2019, first deploying a prototype robot in Cyberport, Hong Kong.
• Today they’ve deployed 400–500 robots across Hong Kong and Japan.
• Their focus: hotel and hospitality robots, delivery, and utility robots in high-rise environments.
• Deployments often require elevator integration and heavy on-site engineering.

Victor summed up the reality of running a robotics business: “You have a perfect hardware, you have a perfect product, but then you also need to work with a ton of third-party partners to make the whole scenario work… You need a lot of field engineers to go on-site and deploy robots.”

This is exactly the pain we’re trying to alleviate: reduce the time, cost, and expert labor needed to deploy robots in the real world by letting them tap into existing domains and maps.

Why Rice Moved In With Us

Rice has now moved into our shared robot lab. Nils says, “Now we’re literally neighbors. There’s a lot of Rice robots all over now.”

Over the coming days and weeks, we’ll be:

• Doing the first Rice x Auki integration
• Making it much easier for venues that already use our network to bring in Rice robots
• Making it easier for us to resell Rice robots into existing Cactus/real world web deployments

Nils: “We will make sure that the people in the Rice community can set up real world web domains, scan their place, connect it to the Rice robot so that the Rice robot knows its way around.”

Hong Kong robotics OGs plus our spatial network is a very natural fit.

Rice x Web3: Floki Bots and the Rise Token

Rice isn’t new to Web3 either. Victor shared how they launched an AI companion robot with Floki:

• Floki is one of the largest meme communities on BNB Chain.
• In May, Rice launched a limited presale of a Floki-themed companion bot.
• They sold 807 units in 24 hours.

This bot acts as an AI companion, performs data collection tasks, and rewards users with $RICE tokens in return.

Aaron added more detail on the upcoming protocol:

• They’re building a protocol on top of the AI companion, with a testnet planned around Q1.
• The idea is to use Auki’s tech to:
  • Help the mini bot get better navigation inside homes
  • Collect anonymized data about house layouts and formats across many homes
  • That data helps train more capable physical AI, while rewarding users who opt in.

From our side: “Physical AI needs to get trained on data about the real world, and Rice is giving the community an opportunity to contribute data and get rewarded.”

Why Hong Kong and the Greater Bay Area Are an Unfair Spawn Point

We also stepped back to talk about where all this is happening.

The Greater Bay Area

We pulled up a map of the Greater Bay Area (GBA), which includes:

• Hong Kong
• Shenzhen
• Guangzhou
• Macau
• Several other cities

Some stats we shared:

• The GBA is smaller than Greater Los Angeles in area
• Yet it holds around 87 million people
• This region has more high-rise buildings than all of North America and Europe combined

It’s also one of the four global hubs for hardware manufacturing and robotics.

As Victor said: “If you’re working in hardware, you want to do something great in robotics, you want to be there. It’s the best place.”

For robotics and physical AI, it’s about as good a spawn point as you can get.

Rice x peaq x Auki: More SDKs, More Interop

We also touched on Rice’s collaboration with peaq:

• Rice plans to integrate the peaq Robotics SDK into their mini bots.
• That means leveraging:
  • peaq's universal machine IDs
  • Their upcoming Get Real campaign
• Once integrated, Rice robots will be able to participate directly in peaq's robotics ecosystem as well.

From our angle, Auki + peaq + Rice is a strong interoperability triangle:

• Our real world web handles spatial perception, mapping, and positioning.
• peaq contributes machine identity and coordination tooling.
• Rice brings deployed physical hardware and Web3-native robots.

We also hinted at another well-known robotics/Web3 project that might be joining our lab soon, but kept that part for off-record discussion.

What’s Next

The main takeaways from this update:

• You can now see very concretely how a brand-new robot can be brought online in minutes using an existing domain on the real world web.
• Rice AI has joined us in the lab, bringing years of deployment experience and hundreds of robots in the field.
• Their Floki companion bot and $RICE token show how companion robots + data collection + Web3 incentives can work in practice.
• The Greater Bay Area remains an absolute powerhouse location for hardware, robotics, and manufacturing, and we’re right in the middle of it.

As always, if you want the unfiltered version, join us live in the Discord. We do the official part on X, then shut off the stream and talk off the record with the people actually in the room.

‍

Watch the whole update on X.

About Auki

Auki is making the physical world accessible to AI by building the real world web: a way for robots and digital devices like smart glasses and phones to browse, navigate, and search physical locations.

70% of the world economy is still tied to physical locations and labor, so making the physical world accessible to AI represents a 3X increase in the TAM of AI in general. Auki's goal is to become the decentralized nervous system of AI in the physical world, providing collaborative spatial reasoning for the next 100bn devices on Earth and beyond.

X | Discord | LinkedIn | YouTube | Whitepaper | auki.com